Project Summary Children with autism spectrum disorder (ASD) are at significant risk for language delay, and roughly 25% of children with an ASD diagnosis will not go on to develop functional speech (Tager-Flusberg & Kasari, 2013). This is concerning, as early language development has long term implications for quality of life in this population (Howlin et al., 2000). Research has suggested that early interactions around toys between children and caregivers play a central role in the development of both expressive and receptive language (Adamson, Bakeman, Deckner, & Romski, 2009). The goal of this project is to examine the developmental associations between a particular type of joint engagement, termed higher order supported joint engagement (HSJE) and vocabulary, when children with ASD are just beginning to speak. This project will use a dataset from a completed NIH-funded project, Predicting Useful Speech in Children with ASD (RO1 DC006893-01A2). Eighty-seven children participated, and were between the age of 24 and 48 months at study entry. The current project will use data collected at two time points, eight months apart. In a previous study in young children with ASD (Bottema-Beutel, Yoder, Hochman, & Watson, 2014), HSJE was found to positively predict expressive vocabulary. Further, when the caregiver talked about the child?s focus of attention (talk that is referred to as follow-in utterances or FI) during HSJE, this form of engagement positively predicted receptive vocabulary. Using a longitudinal, cross-lagged panel design, the proposed project will extend this research by providing evidence about the directionality of these effects (Raghunathan, Rosenthal, & Rubin, 1996). This approach compares the magnitude of the correlation between two longitudinal, cross-lagged variables (e.g., early expressive vocabulary and later HSJE vs. early HSJE and later expressive vocabulary), while accounting for concurrent correlations between the two variables and the temporal stability of the variable across the two time points. If the correlation in one direction (e.g., early expressive vocabulary and later HSJE) is larger than the correlation in the opposite direction (e.g., early HSJE and later expressive vocabulary), this provides evidence that the early variable in the larger correlation is ?driving? the later variable. It is expected that early expressive vocabulary will be found to ?drive? later HSJE, and early HSJE + FI will be found to ?drive? later receptive vocabulary. Findings from this study will aid in the development of developmentally-based interventions that use joint engagement as a format for achieving developmentally distal intervention outcomes. Constructs that are found to ?drive? development (e.g., expressive vocabulary and HSJE + FI) can be leveraged as primary targets of intervention, with the expectation that they will bootstrap the development of later outcomes (e.g., HSJE and receptive vocabulary) even after the intervention has stopped.